Isotopic ion flux assays are a proven and classical (over 25 years of use) means to characterize function of voltage- or ligand-gated ion channels in clonal cell lines.3536 Ion flux assays complement, and in many ways offer advantages over, more tedious electrophysiological analyses of channel function. Ion flux assays adapted for stopped-flow studies like those used in enzymology give temporal resolution comparable to that for the fastest electrophysiological studies and are fully suitable for intricate study of nAChR functional kinetics.37 Membrane vesicles very rich in nAChR or other channels of interest and resistant to fluid pressures attained in stopped-flow studies, but not to whole cells, are preparations of choice for such studies. Nevertheless, ion flux assays using intact cells are ideally suited for high-throughput analyses of nAChR function using very simple techniques and common instrumentation. Ion flux assays integrate responses for the ensemble of nAChR from the entire population of cells in a cell culture dish or microwell, typically summed across >108 receptors (>105 cells per 15.5-mm diameter well (24-well tray) with ~103 surface receptors per cell).3839 Cellular nAChR responses can be determined using ion flux assays with a temporal resolution of seconds, and ion flux rates typically remain constant over a period of 45 to 60 seconds extrapolated through zero ion flux at time zero. When bi- or multiphasic kinetics of ion flux have been observed, it has reflected time-dependent inactivation of nAChR rather than exhaustion of accessible radiotracer.
Agonist dose-response profiles for test ligands can be obtained from studies of cells plated at equal density in wells in a multiwell array each incubated with a different dose of the test ligand. Positive control (total) responses are determined in samples exposed to a maximally efficacious dose of a standard agonist. Nonspecific ion flux is determined as a negative control in samples lacking agonist or containing both agonist and fully-blocking antagonist. Nonspecific ion flux is subtracted from positive control or test sample responses to yield specific ion flux for those samples. Specific ion flux is plotted as a function of concentration of test ligand, and the data are analyzed using nonlinear regression fits to the formula F = Fmax / (1 + (EC50/[L])n), where F is the measured specific ion flux and [L] is the molar ligand concentration, to yield the parameters Fmax as the maximum ion flux, EC50 as the ligand concentration giving one-half of the maximal ion flux response to the standard agonist, and n (> 0) as the Hill coefficient for the process. Efficacy of test ligand can be determined relative to maximal response to standard agonist in positive control samples. Some test ligands may produce maximal responses at high concentrations that are less than the maximum response to the standard agonist. If agonist dose-response curves for these ligands plateau, then they act as partial agonists and their potencies can be estimated by the concentration giving one-half of their maximal effect. If agonist dose-response curves are bell-shaped, then the test ligand may be exhibiting self-inhibition of functional responses and/or inducing desensitization of nAChR at higher concentrations, but its potency also can be estimated as the dose giving one-half of its maximal effect. EC50 values also will differ from the concentration of a test ligand giving one-half of its maximal effect for the super efficacious drug having >100% of standard agonist efficacy. To provide valid measures of agonist activity, dose-response curves should include measurements at agonist concentrations at least ten times higher than the apparent EC50, and responses at those concentrations should not be more than double responses obtained at the apparent EC50.
Antagonist dose-response curves for test ligands can be obtained from studies of samples treated with different doses of the test ligand in the presence of a standard agonist at a constant concentration. Specific ion flux results are plotted as a function of concentration of test ligand, and the data are analyzed using the formula F = Fmax / (1 + (IC50/[L])n), where F is the measured specific ion flux and [L] is the molar ligand concentration, to yield the parameters Fmax as the maximum ion flux in the absence of antagonist, IC50 as the ligand concentration giving a half-maximal inhibition of ion flux response, and n as the Hill coefficient for the process, which is < 0 for an antagonist in this formula. Tentatively, affinity of nAChR for the test ligand can be expressed as IC50 value (see below). Even if a ligand fails to produce blockade to negative control levels of ion flux, its affinity for nAChR can be tentatively expressed as the concentration giving one-half of that ligand's maximal degree of block.
Competitive or noncompetitive mechanisms of functional blockade can be distinguished based on studies of agonist dose-response profiles at zero and fixed antagonist concentrations. These curves shift to the right (to higher observed EC50 values, thus showing that functional block is surmountable) as competitive antagonist concentrations increase, but they shift downward (reflecting diminished agonist apparent efficacy in the face of insurmountable block) without substantially changing observed EC50 values as noncompetitive antagonist concentrations increase. Similarly, competitive antagonist dose-response profiles will shift to the right (giving increases in apparent IC50 values) as agonist concentrations within the maximally efficacious dose range increase, but noncompetitive antagonist dose-response curves will not shift appreciably left or right as agonist concentrations vary within the maximally efficacious range. IC50 values for noncompetitive antagonists are equal to K values (measures of functional nAChR affinity for the ligand; concentration at which there is half-maximal occupancy of nAChR) regardless of agonist concentration used (although agonist concentration should be equal to or greater than its EC50 value for practical reasons). However, determination of Ki values for competitive antagonists requires additional analysis. The competitive antagonist Ki value can be estimated as the concentration of antagonist that produces a doubling in apparent EC50 value for an agonist in an agonist dose-response profile relative to the EC50 value obtained from such a profile in the absence of antagonist. Competitive antagonist Ki values can be determined more precisely from nonlinear regression analysis of agonist dose-response curves and an expression describing receptor occupancy by agonist and competitive antagonist.40
Ion flux assays can also be used to derive additional information. Use dependence of blockade can be evaluated by testing for enhanced antagonism after short pre-treatment with agonist. Insights into voltage sensitivity of functional blockade can be gained by studies of ion flux responses in the presence of extracellular medium containing different concentrations of potassium ion. Extracellular ion substitution experiments (e.g., N-methyl-D-glucamine exchanged for sodium, Ca2+ removal) can give insights into ion selectivity of channels under study. Spontaneous opening of channels can be assessed by comparing levels of ion flux in the absence of agonist to flux in samples treated with antagonist alone.
On balance, ion flux assays can give information comparable and complementary to that obtained from whole cell current and other methods of electrophysiological recording. Single channel analyses remain the purview of electrical recording. Whole cell current recording has advantages in studies of acute desensitization (occurring in seconds or less) and in some studies of very rapidly inactivating channels that are not open long enough to give significant, integrated ion flux signal above background, such as a7-nAChR. Theoretically, agonist dose-response profiles obtained from peak whole cell currents might differ from those obtained at later times in whole cell current records or from integrated ion flux responses if rates of desensitization of a given nAChR subtype differ across agonists. Similarly, dose-response profiles for a given agonist acting at different nAChR subtypes might differ when taken from peak whole cell currents or when derived from integrated ion flux responses if the two nAChR subtypes have different rates of desensitization. However, there is no reason why antagonist dose-response profiles determined at agonist EC50 values should differ when using, for example, peak whole cell current or integrating ion flux assays. In practice, very few studies have made direct comparisons between ion flux and whole cell current results. Experience shows that dose-response profiles differ more for analysis of a given nAChR subtype expressed in different systems (mammalian cell vs. Xenopus oocyte) than when analyzed using ion flux and electrophysiological techniques in the same expression system.
86Rb+ efflux assays have the highest sensitivity and resolution of any isotopic ion flux assay of nAChR function tested.38, 41 They can be applied to nAChR because all indications are that nAChR channels are large and relatively nonspecific for size of monovalent cations. Even though the physiologically important current mediated by most nAChR is an inward current carried by Na+, monovalent cations will flow down their concentration gradients, into or out of the cell, when nAChR are opened. 86Rb+ efflux assays essentially establish an infinite gradient for 86Rb+, which rapidly leaves the cell on nAChR stimulation. The specificity of 86Rb+ efflux assays for nAChR responses has been demonstrated repeatedly. Nevertheless, pharmacological studies should be done to discount potential contributions of voltage- or Ca2+-gated K+ channels to nicotinic agonist-triggered 86Rb+ efflux responses in a specific cell type. Note, however, that many different kinds of channel blockers, including voltage-gated K+ channel blockers such as tetraethylammonium and 4-aminopyridine, voltage-gated Ca2+ channel blockers such as dihydropyridines, ionotropic glutamate receptor blockers such as MK-801, etc., also block nAChR channels, sometimes with comparable affinities.
In a typical protocol, cells of interest harvested from master plates via trypsiniza-tion (see above) are seeded into wells of multiwell trays and maintained overnight in the incubator. Cells are typically seeded in 0.5 to 1 ml of growth medium to achieve confluence at the time of assay (2-5 • 105 cells per 15.5-mm diameter well in a 24-well tray; 150 to 300 |g of total cell protein per well, depending on cell type). If cell plating density on the planned day of assay is too low, then ion flux assay signal will be suboptimal, and delaying the assay until cells achieve confluence is recommended. If cells are overconfluent, then there is a risk that cells will lift as a sheet during sample processing, most notably partially around the edges of the plate, causing loss of cells in initial rinses and/or transfer of cells to the efflux sample. Enhanced attachment of some cell types can be effected using specially treated plates (e.g., Falcon Primaria plates for TE671/RD cells). A more standard routine to prevent cell lifting during the assay is to treat any kind of plate momentarily with 100 |g/ml poly-D-lysine (70,000 to 150,000 Da average size; 1 ml per 15.5-mm diameter well) before seeding cells. Function of some nAChR subtypes is compromised when cells expressing those nAChR are seeded onto plates treated with polyethyleneimine, so pilot studies are recommended to ensure that poly-D-lysine, polyethyleneimine, or polyornithine treatment of assay plates does not cause inhibition of nAChR function. Note that cells will be rinsed several times before assay of nAChR function commences, so any effects of residual fluid phase attachment factor should be minimized in most instances; plates are not routinely rinsed after poly-D-lysine treatment before seeding cells. For some cell types (e.g., BC3H-1), cell lifting occurs unless cells are seeded into 35-mm diameter dishes or into wells of 6-well trays. Probably because of the sheet-like nature of growth of those cells, they lift easily from smaller diameter wells that have larger ratios of circumference to surface area. For cell types that tend to clump rather than adhere to substratum, such as PC12, SH-SY5Y or IMR-32 cells, creating a fine, single-cell suspension of adequate concentration to ensure broad and uniform dispersal of seeded cells gives optimal results. For cell types like IMR-32 cells that tend to lift even when seeded appropriately and onto a good substratum, placement of multiwell trays on a warming plate at 37oC and use of warm instead of room-temperature rinsing media during processing (see below) helps to prevent cell lifting.
Plated cells are loaded with 86Rb+ by aspirating seeding medium (typically serum-supplemented DMEM) and replacing it with otherwise identical medium typically supplemented with 1 to 2 |Ci/ml of isotope (250 |l per 15.5-mm diameter well). Because 86Rb+ has such a short half-life, preparation of loading medium from isotope stocks to achieve such a level of activity is empirical, and the amount of 86Rb+ in loading medium for each experiment should be noted. Obviously, the amount of 86Rb+ in loading medium can be decreased to conserve resources if the nAChR function being assessed is spectacular or can be increased to enhance signal in studies of nAChR with low functional activity and/or if cells are expressing low numbers of functional nAChR. Samples are returned to the incubator for at least 4 hours to allow activity of the Na-K-ATPase to concentrate 86Rb+ (as a K+ analogue) in intracellular medium. Studies of the kinetics of 86Rb+ loading should be done for every cell type being assayed, but experience indicates that loading is >90% complete over 4 hours. It is recommended that some cell samples be used to determine the amount of 86Rb+ loaded and the amount of 86Rb+ remaining in the extracellular fluid. Specific radioactivity of 86Rb+ can be calculated from the latter value and from knowing the extracellular volume and the concentration of extracellular K+ and analogues. Based on the calculated specific activity of 86Rb+, the determined amount of 86Rb+ loaded into cells, and an assumed or determined intracellular concentration of K+ and analogues (typically ~120 mM), intracellular volume accessible to 86Rb+ can be calculated.
Once loaded with 86Rb+, plated cells are removed from the incubator, and subsequent procedures are conducted behind a Lucite shield at room temperature (with exceptions noted above), in part to slow functional desensitization. Loading is terminated by aspiration of medium into a shielded collection flask and application of ion efflux buffer (130 mM NaCl, 5.4 mM KCl, 2 mM CaCl2, 5 mM glucose, 50 mM HEPES, pH 7.4, ~300 milliosmolar; typically 3, 2, or 1 ml, respectively, per well for 12-, 24-, or 48-well trays). In some early studies, cell culture medium was used as the efflux buffer. However, phenol red can have activity as an nAChR antagonist, and serum components can include esterases that can cleave ester bond-containing nicotinic ligands like acetylcholine, suberyldicholine, or succinylcholine, so it is better to avoid using medium and serum in efflux assays. Buffer aspiration and cell rinsing are repeated two more times for a total time of 20 seconds to 2 minutes. Rinse buffer is then replaced with fresh ion efflux buffer usually containing nicotinic ligands of choice. Whereas stock solutions of efflux buffer are made up and stored at 4oC until use, drugs are made fresh daily from powder unless control studies show that frozen aqueous stocks or stocks dissolved in other media (e.g., ethanol or dimethylsulfoxide) give the same results as freshly prepared ligand. An option is to create high concentration stock solutions of more expensive ligands sold in small quantities or of commonly used ligands. Stocks of 1 M carbamylcholine in water, 10 mM d-tubocurarine in efflux buffer, 1 M nicotine in water, and 100 mM methyllycaconitine in ethanol can be stored at -20°C without loss of drug activity. After incubation for a prescribed period (1 to 5 minutes for studies in the laboratory), efflux medium is collected for counting, and cells are dissolved in 0.01 N NaOH, 0.1% sodium dodecyl sulfate for further analysis. 86Rb+ can be quantified by Cer-enkov counting of aqueous samples in liquid scintillation counters at ~25%/45% efficiency (glass/plastic vials) or by scintillation counting in scintillation fluid at ~95% efficiency.
22Na+ influx assays of nAChR (and other channel) function have been described in detail elsewhere and will not be repeated here.35 36 86Rb+ influx assays, designed as variations on methods established for 22Na+ influx assays, also have seen substantial use in studies of nAChR.42-44 Cells are plated and processed and solution volumes are used as for 86Rb+ efflux assays. Sample processing is initiated by removal of cell growth medium by rinsing cells and incubation for up to 30 min in rinse/equilibration medium. For cell rinses and equilibration, some published methods used complete or serum-free cell culture medium (~330 milliosmolar),41,44 sometimes because cells were preincubated for long times with nicotinic ligands. For simpler pharmacological characterization, HEPES-buffered salt solutions (e.g., 150 mM NaCl, 5 mM KCl, 1.8 mM CaCl2, 1.2 mM MgCl2, 0.8 mM NaH2PO4, 10 mM glucose, 15 mM HEPES, adjusted to pH 7.4 with NaOH, ~310 milliosmolar)42 are used for rinses and equilibration. For ion flux assays, Na+-free influx buffer (replacing NaCl in whole or in part with sucrose, e.g., 0.25 M sucrose, 5 mM KCl, 1.8 mM CaCl2, 10 mM glucose, 15 mM HEPES, pH 7.4, ~325 milliosmolar)44 is substituted. Use of Na+-free medium for influx assay reduces extracellular monovalent cations that could compete for influx of radiotracer and minimizes changes in membrane potential during nAChR activation. Just before starting the influx assay, all samples are subjected to 1 minute of incubation in Na+-free influx buffer supplemented with 1 to 2 mM ouabain to inhibit action of the Na+-K+-ATPase. To initiate influx, this medium is then simply removed by aspiration, and cells are bathed in influx buffer supplemented with 1 mM ouabain (to continue inhibition of 86Rb+ uptake via Na-K-ATPase), 86Rb+ (about 5 ^Ci/ml), and nicotinic ligands of choice. After a prescribed period of time (20 sec to 2 min for studies in the laboratory), the influx period is terminated as cells are rinsed 3 to 4 times over ~0.5 to 3 minutes to remove extracellular 86Rb+. 86Rb+ uptake is quantified by Cerenkov or liquid scintillation counting of cellular samples dissolved in 0.01 N NaOH, 0.1% sodium dodecyl sulfate.
86Rb+ or 22Na+ influx assays can and do provide essentially equivalent information to that obtained through 86Rb+ efflux assays. Influx assays have advantages for some types of studies. For example, studies of nAChR desensitization involving pretreat-ment of cells with nicotinic ligands are simpler to interpret initially if influx assays are used to monitor nAChR function. Use of 86Rb+ efflux assays for such studies is potentially complicated if ligand pretreatment causes transient activation of nAChR and loss of loaded 86Rb+ from cells before challenge doses of ligand are applied to initiate nAChR functional analysis. However, it has been demonstrated that normalization of 86Rb+ efflux data to the amount present in cells at the time of efflux assay initiation, even when used in studies involving pretreatment of cells with nicotinic ligands, adequately accounts for any loss of loaded 86Rb+. 86Rb+ efflux assays are superior to either kind of influx assay in terms of resolution (signal:noise) and sensitivity (signal when using the same amount of isotope and/or biological material), which translates into lower cost. Moreover, emissions from 86Rb+ are less energetic than those from 22Na+, which translates into improved safety. In addition, 86Rb+ can be detected using Cerenkov counting without use of scintillation vials or fluid, again making them advantageous economically and in terms of safety (no need to purchase and dispose of organic scintillants). On balance, 86Rb+ efflux assays are the clear choice for ion flux analyses of nAChR function.
Sample handling in isotopic ion flux assays most commonly involves using conventional or (for some steps in the method) repeating pipettes for solution application or aspiration. In typical 86Rb+ efflux assays, the sample processing interval for a 24-well tray is 12 seconds, and sample wells are processed sequentially one at a time. For a typical efflux period of 5 minutes, 10 minutes pass while processing a single plate. Plates are tilted during solution exchange so that fluids are removed or applied more gently via laminar flow to the cell plating surface. Pipette tips for solution aspiration and application also are positioned consistently at one point inside the bottom edge of the well, and care must be taken to prevent or minimize displacement of some cells or a spot of cells to minimize data scatter.
A recently developed novel approach (the "flip-plate" method) for conduct of ion flux assays provides higher throughput, gentler sample handling, more flexibility in experimental design, superior temporal accuracy, and improved sample-to-sample reproducibility than sequential pipetting methods. Cells are seeded into poly-D-lysine-coated wells and loaded with 86Rb+ (in this example for efflux assays) as usual on Falcon or Corning multiwell plates. Only Falcon or Corning brands of multiwell plates are currently manufactured so that the lips at the top of each well are elevated/ level relative to the lips at the edge of the plate. This allows formation of a tight seal between well lips when two plates are opposed top to top and pressed together under uniform finger pressure. 86Rb+ loading medium is aspirated from wells of "cell plates" as usual into a shielded collection flask, but then the cell plate is inverted and aligned top to top over the first of two "rinse plates" set up in advance to contain fresh efflux buffer (2 ml per 15.5-mm diameter well). The plates are held together firmly, and the ensemble is gently and slowly flipped (rotated) so that rinse buffer bathes the cells. After a few seconds, the plates are then gently flipped back to allow the rinse solution to fall back into the first rinse plate. The cell plate is lifted off, inverted over the second rinse plate, and rinsed in the same fashion. (Rinse plates can be washed and reused; the bulk of isotope is safely removed during the initial aspiration step.) There is a small, consistent, residual volume of buffer that remains in the cell plate (typically 30 to 40 ||L for a 24-well plate) after removal of extracellular 86Rb+ and rinses. Therefore, to maintain exact test concentrations, the residual rinse solution is aspirated from each well as quickly as possible (usually about 10 seconds total time per plate). The cell plate is then inverted and aligned over an "efflux plate" set up in advance to contain 2 ml of nicotinic ligands of choice in efflux buffer. Efflux is initiated by again holding the plates firmly together and gently flipping them, allowing the test concentrations of drug to fall into and bathe cells in wells of the cell plate. After a prescribed incubation period, the plates are again held firmly together (sometimes gently swirled once to displace pericellular 86Rb+) and gently flipped, allowing assay solution containing effluxed 86Rb+ to fall back into the efflux plate. Although a small amount of residual volume remains in the wells of the cell plate, this quantity is uniform and can be determined to allow data normalization. The wells of the cell plate are then filled with the same volume of solution as was used in the efflux plate but containing 0.01 N NaOH, 0.1% sodium dodecyl sulfate to dissolve the cells and their contents. Samples from each well of the efflux plate and the cell plate can then be used to determine levels of effluxed and remaining intracellular isotope, respectively, for each sample after transfer to scintillation vials for Cerenkov or liquid scintillation counting. It has been found to be more convenient, more economical, and less labor intensive simply to prepare the plates for Cerenkov counting using an E&G Wallac (now Perkin-Elmer) Tri-Lux 1450 Microbeta plate-reading liquid scintillation and luminescence detector. The flip-plate method has been used routinely in 6-, 12-, 24- and 48-well formats. Use of a 96-well format is, however, not recommended, as surface tension and the requirement for higher tolerance in well dimensions and orientation make assays problematic. The flip-plate method has also been used successfully after appropriate adaptation for influx assays. Variants on the flip-plate approach have potential applicability in virtually any assay using cells or substrate adherent to or immobilized on a test plate.
Was this article helpful?